File size: 2,519 Bytes
d98ba57
 
cc2ce8c
 
 
 
 
 
d98ba57
cc2ce8c
 
 
d6936f0
cc2ce8c
d6936f0
cc2ce8c
d6936f0
cc2ce8c
 
 
 
d98ba57
cc2ce8c
d6936f0
cc2ce8c
 
 
 
d98ba57
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cc2ce8c
 
 
 
 
d6936f0
cc2ce8c
 
 
d98ba57
 
 
 
528bf3d
 
d98ba57
 
 
 
 
cc2ce8c
 
 
d98ba57
cc2ce8c
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
import os

from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.vectorstores import Chroma
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.document_loaders import PyPDFLoader

from .embeddings import EMBEDDING_MODEL_NAME
from .vectorstore import PERSIST_DIRECTORY, get_vectorstore


def load_data():
    print("Loading data...")
    docs = parse_data()
    print("Loaded documents")
    embedding_function = HuggingFaceEmbeddings(model_name=EMBEDDING_MODEL_NAME)
    print("Building index...")
    vectorstore = get_vectorstore(embedding_function)

    assert isinstance(vectorstore, Chroma)
    vectorstore.from_documents(
        docs, embedding_function, persist_directory=PERSIST_DIRECTORY
    )
    print("Index built")
    return vectorstore


def parse_data():
    docs = []
    for root, dirs, files in os.walk("data"):
        for file in files:
            if file.endswith(".pdf"):
                file_path = os.path.join(root, file)
                loader = PyPDFLoader(file_path)
                pages = loader.load_and_split()

                # split it into chunks
                text_splitter = RecursiveCharacterTextSplitter(
                    chunk_size=1000, chunk_overlap=0
                )
                doc_chunks = text_splitter.split_documents(pages)

                for chunk in doc_chunks:
                    chunk.metadata["name"] = parse_name(chunk.metadata["source"])
                    chunk.metadata["domain"] = parse_domain(chunk.metadata["source"])
                    chunk.metadata["page_number"] = chunk.metadata["page"]
                    chunk.metadata["short_name"] = chunk.metadata["name"]
                    docs.append(chunk)

    return docs


def parse_name(source: str) -> str:
    return source.split("/")[-1].split(".")[0].replace("_", " ")


def parse_domain(source: str) -> str:
    return source.split("/")[1]


def clear_index():
    for filename in os.listdir("../chroma_db"):
        file_path = os.path.join("../chroma_db", filename)
        try:
            if os.path.isfile(file_path) or os.path.islink(file_path):
                os.unlink(file_path)
        except Exception as e:
            print("Failed to delete %s. Reason: %s" % (file_path, e))


if __name__ == "__main__":
    clear_index()
    db = load_data()
    # query it
    query = (
        "He who can bear the misfortune of a nation is called the ruler of the world."
    )
    docs = db.similarity_search(query)
    print(docs)